Text Analysis of the Social Capital Literature Network ... - INSNA

38 Building a Theory of Social Capital / Lin
Another prevailing measurement strategy focuses on network locations as measurements of
social capital . A major perspective is the argument that bridges or access to bridges facilitates
returns in actions . Granovetter's notion of bridges as expressed in the strength of weak ties
(1973) was a preview of this argument, which is elaborated and formalized by Burt in his
notions of structural holes and constraints (1992) . Other measures of bridges (e .g ., betweenness)
would also be candidates for social capital, even though they are used less in the social capital
context .
There are many other measures such as size, density, cohesion, and closeness of social networks
which are candidates as measures for social capital . However, research evidence is much less
clear as to their viability in a social capital theory . Unless clear theoretical arguments are
presented along with the use of any specific measures, as both measures of social resources and
network locations have been, it would be ill-advised to simply use any network measure as an
indicator of social capital .
Sampling Techniques
Three sampling techniques have been employed to construct measures of social capital, as can
be seen in Table 4 . The saturation sampling technique is useful when it is possible to map a
definable social network . In such networks, data from all nodes are gathered and their
relationships identified, and measurements of network locations can be developed . The
advantage of this technique is that it allows detailed and complete analyses of each and every
network location as well as embedded resources in each node . Because of the requirement that
the network has a defined and manageable boundary, it is a technique most useful for studies
of social capital within an organization or a small network among organizations .
Table 4 . Measurement Techniques
Technique Advantages Disadvantages
Saturation
survey
Name generator
Position
I generator
Complete mapping of network
Customized content areas
Ego-centered network mapping
Content-free
Sampling of hierarchical positions
Multiple "resources" mapped
Direct and indirect accesses
Limited to small networks
Lack of sampling frame
Biased toward strong ties
Lack of specificity of
relations
For larger and less definable networks, ego-network sampling techniques are used . Typically,
the name-generator technique (Laumann, 1966 ; Wellman, 1979 ; McCallister & Fischer, 1978 ;
Burt, 1984; Marsden, 1987) is employed. This measurement technique elicits a list of ties from
ego, and the relationships between them as well as among them are identified . From these data,
locations of ego as well as these ties, relative to one another, can be computed . Network
resources can also be obtained from the name-generator technique . Measures such as
composition (typical resource characteristics), heterogeneity (diversity of resources), and upper
reachability (best possible resources) can be computed . The advantages of this approach
include (1) the identification of specific content areas, relative to actions under investigations,